Building Apps with AI Tools: ChatGPT, Semantic Kernel, and Langchain

Introduction

• Using GPT to build software
• Setting up your environment

1. Connecting to the ChatGPT API

• Introduction to ChatGPT and its parameters
• Connecting to the ChatGPT API

2. A Discount Bot

• Sentiment analysis
• Providing a discount through a UI
• Challenge: Turning rude customers away
• Solution: Turning rude customers away

3. Building with Semantic Kernel

• A simple summarizer with Semantic Kernel
• Formatting your data with few-shot learning
• Integrating chain-of-thought reasoning into your app
• Learning to use Whisper for text to speech
• Challenge: Creating a "thinking out loud" librarian
• Solution: Creating a "thinking out loud" librarian

4. Building Document Search with LangChain

• Building a simple prompt chain with LangChain
• Answering questions with a vector DB
• Semantic search and embeddings
• Extracting key information from your question
• Challenge: A librarian with a library
• Solution: A librarian with a library

5. Testing ChatGPT Apps

• Generating sample data with ChatGPT
• Generative AI–powered tests
• Evaluating GenAI prompt performance
• LLM framework security
• Challenge: Building a GenAI test suite for your librarian
• Solution: Building a GenAI test suite for your librarian

Conclusion

• Next steps